It's interesting to think how his career would look if he had come up to the majors as a pitcher instead of an infielder.

Am I the only voter strongly considering putting Walters pretty high on my ballot?

I'm looking forward to seeing what Chris J. has to say about Walters, but at first glance, and in my system, even with war discounts, this guy's got a monster peak, and a better peak IMO than Ferrell. I see Walters, initially, as placing at or near Mendez's slot on my ballot.

Bill James wrote in his book on managers that Bill McKechnie stacked his Reds with excellent defense which helped the Reds get good performances out of all of their pitchers. Do any of the advanced metrics prove or disprove this hypothesis?

look at Bucky's BP page,
http://www.baseballprospectus.com/dt/waltebu01.shtml
and see how his DERA (adj for defense) comps to his NRA; and how it changes remarkably going form PHI to CIN. Yes, his defense did help. Yes, even accounting for this, his WARP still comes out very good. He'd make my prelim ballot.

I have always said that there is a case to be made for Bucky Walters as the next Golden Age pitcher and better than some who are already in the HoF. But that is assuming a lower standard than what the HoM is arriving at, what with more 19C and NeL HoMers than HoFers and therefore fewer 20C MLers.

As it is, it is hard to go too crazy for a guy who stacks up as follows:

Career

Walters 115 ERA+ in 3104 IP
Pennock 106 in 3572
Hoyt 112 in 3762
French 114 in 3152
Root 110 in 3197
Harder 113 in 3426
Mays 119 in 3021
Luque 117 in 3220
D. Leonard II 119 in 3218
Shawkey 114 in 2937

But of course his case is a peak/prime case not a career case:

Walters 124 in 2634 in 10 years
Passeau 120 in 2523 in 10
Rixey 122 in 2633 in 11
Faber 128 in 2099 in 8
Ruffing 127 in 2017 in 8
Ferrell 128 in 2256 in 8
Hoyt 120 in 2025 in 8

Granted that's a lot better. But the guys we've elected from the Golden Age, other than Faber, have peak/prime ERA+ of about 167 (Grove), 148 (Alex and Hubbell), 135 (Vance). And some other candidates are at 134 (Bridges), 138 (Gomez), 141 (Luque), 136 (Mays), 143 (Warneke), Dean (133), Shocker (131), Harder (135), Pennock (133). Granted all but Bridges did it in under 2000 IP. Bridges at 134 in 2284 IP looks pretty good.

But from other eras we still have Bond (130 in 2865), McCormick (128 in 2689), Welch (126 in 3121), Griffith (135 in 2221‚ (Joss (148 in 2220), Cicotte (129 in 2905), even Babe Adams (131 in 2237). Granted those were all different times.

Put it all together, anyway, and I've got Bucky in my top 20 but trailing (through eligible 1959, and of course always subject to change):

Waddell
Newhouser
Bond
Joss
(Lyons)
Dean
Ruffing
Rixey
McCormick
Cicotte

Griffith
Gomez
(Faber)
W. Cooper
Ferrell
Walters

TomH,

Thanks for the link.

Walters gets a lot of little credits. He's a good hitter for a pitcher with a .226 EQA. He's a good fielder for a pitcher at 20 RAA. His big year is huge in WARP1 at 13.1...but it improves to 13.2 after going through the WARP3 tilt-a-whirl.

For the peak/prime voters out there...

WARP1, best to worst 10 seasons
13.1
9.5
9.2
8.6
6.6
6.1
5.6
4.7
4.3
3.9
====
75.4

WARP3 10 best seasons
13.2
9.8
9.0
8.4
8.2
7.5
5.7
5.4
4.8
4.3
====
76

I'm not well versed in the WARPs, but that looks like a peak/prime candidate to me.

Compare to current peak/primer darling Wes Ferrell.

Ferrell's best 10 by WARP1 looks great
14.1
11.9
11.4
10.7
9.7
8.9
6.6
6.5
5.1
1.8
====
86.9

but he gives back a lot of in WARP3
13.5
10.4
9.9
9.7
8.8
8.1
5.9
5.5
5.4
2.2
====
79.3

Walters out WARPs Ferrell over the remainder of their careers

Ferrell 1.2 WARP1, 1.6 WARP3

Walters 13.4 WARP1, 13.1 WARP3

So it's kind of an interesting comparison because of the dilution of value from Ferrell's WARP1 to his WARP3. If you use or value WARP3, you're bound to find these guys extremely comparable, and there's solid rationales available for voting one above the other and vise versa.

Win Shares also sees them as fundamentally similar but with the same higher concentration of value into fewer seasons for Ferrell. Here's their top-ten WS seasons:

BW   WF
==   ==
38   35
32   32
32   28
27   27
20   26
16   25
15   18
14   18
14   14
12   6

CAREER
BW   WF
==   ==
258  233

Complicating all this is that, according to Chris J., it appears that Walters's usage pattern was more difficult than Ferrell's. Chris indicates that Walters squared off against the best teams as much as almost anyone he (Chris)has studied, and that Walters rose to the occassion. In addition, Walters's RSI is about 100 (on a 69 career OPS+), but it doesn't remove his own hitting which was substantial. Ferrell's is 102 (on a 100 career OPS+).

Anyway, this one ought to be tricky, but Walters, like Ferrell, is very much on my ballot as of now.

BTW: I don't know much about his repertoire, but he must have thrown a very heavy ball beause his K and BB rates do not suggest the kind of dominance he exhibited.

Bill James has a very detailed write-up on Walters in tne NBJHBA, and he describes Walters as highly similar to Bob Lemon in style. To sum up:

1) Converted third baseman
2) Threw hard, but not a strikeout pitcher
3) Primary pitch the slider
4) poor strikeout/walk ratio
5) great ground-ball fielding support
6) outstanding hitters (for pitchers)

More complications with Walters are any potential war discount for 1943-45. I'm surprised he did so well in the WARP1->WARP3 conversion because of this.

I've got the write-up of Walters up on my site. I'd link to it but this computer that I'm at seems to hate both opening pages as new windows, and the Control C/V features so I can't link it.

Short version: his RSI was league average, but that' s misleading. His MOWPs were farkin' fantastic, though.

Right now I think that Walters is on my ballot, edging out Dizzy Dean. I have him below Ferrel due to Ferrell's hitting. walters may not have been a bad hitting pitcher, but Ferrell was a league average hitter and I think it gives him a slight edge overall. Walters was probably the better pitcher though.

Don't have much to say, but will be on my ballot even with a slight deduction for WWII performance. Have Ferrell slightly ahead.

He's not Ferrell, but I have him slightly below Ruffing. More peak (actually, really just one peak year, but that's more than Red), less career, needs some WW2 discounting. Not sure on an exact spot for him yet, but he will be on ballot.

I see three peak years, at least by WS, where he's got three 30+ WS seasons.

I ran Walters through the RA+ Pythpat machine. The raw numbers came out as 195-150. But some of that was during the war years - you particularly need to deflate his 1944. Making adjustments for the war years by tweaking the park factors, he loses 3 wins, to 192-153. But then we make the same offensive adjustmens as for Ferrell and Ruffing, and he gains back 5 wins, to 197-148. That offensive adjustment does another thing for Walters: becuase his best hitting years coincided with his best pitching years, making the adjustment raises the height of his peak.

Here are my equivalent records for Walters and Ferrell, including all of these adjustments, for offense and war dilution. I've sorted the years from best to worst by equivalent FWP. (Note: the equivalent FWP are computed before the wins and losses were rounded to integrer values.)

Ferrell        Walters
 W-L   FWP      W-L   FWP
24-11   30     26-10   35
23-10   28     23-11   28
21-10   25     21-13   20
22-12   23     20-12   20
21-11   22     11- 7   11
17-10   17     15-13   11
13- 7   15     15-13   10
15-16    7     14-14    7
 1- 0    2      9- 8    7
 1- 1    1     14-14    7
 0- 0    0      1- 0    1
 0- 0    0      0- 0    0
 7- 12  -2      1- 3   -1
                5- 9   -2

In comparing Walters and Ferrell, it appears that Walters had the best single year (1939) but outside of that one year, Ferrell had the slightly peak. Walters has a little more bulk as an ordinary pitcher.Walters's career as an infielder doesn't amount to much, and he doesn't appear to be as valuable a pinch hitter as either Ferrell or Ruffing.

There are still two adjustable issues not yet accounted for here - defensive support, and quality of opposition. Chris J. has information on both of those.

OCF's analysis shows pretty clearly, it seems to me, that Ferrell is the stronger of the two candidates, since he is already looks better here and figuring in defensive support will strongly favor him over Walters.

However, that leaves the quality of opposition element unexamined. There, Walters has a decided edge. While I find it highly unlikely that this edge would be enough to move Walters ahead of Ferrell in my rankings, it could definitely affect his placement relative to other borderline pitching candidates.

But I have no idea of how to include quality of opposition in a systematic way in my rankings.

Is anyone doing this? If so, how?

Or is this something that everyone is just doing informally??

How much of Walters's defensive support is of his doing? He was a slightly below average 3B, but he was well above average with the glove as a pitcher.

Well, since win shares doesn't study pitcher fielding at all, it has no independent role in Chris J.'s metric.

A "thirty seconds of research" study of WARP's use of pitcher fielding suggests that Walters' fielding contributions matter, but not very much.

In 1939, Cincinnati was a great fielding team, 273 FRAA. Pitchers collectively were -8 FRAA, so the other fielders were 281 FRAA. Walters threw 319 out of 1403.7 team innings, so, if his team fielding support was average, the defense would save 64 RAA on Walters' behalf. Walters himself was 3 FRAA, so we can estimate that the team saved 67 FRAA while Walters was pitching. Walters' share is 4.5%.

It's not a meaningless amount: I'd guess it gains Walters a win or two over the course of his career, but it accounts for only a little of the above-average fielding support he received.

SABR Bio Project entry for Bucky Walters

New bio-project entry, interesting stuff.

If I'm using runs allowed, why does pitcher fielding matter . . . it's already in there right?

And how much impact can one pitcher, in at most 1/4 of his teams innings; have on his team's fielding adjustment?

If I'm using runs allowed, why does pitcher fielding matter . . . it's already in there right?

Correct. You have to do this at a team level, and as Joe suggested, any one pitcher won't move the whole-team whole-season stats by very much at all.